A concurrent-chain schedule was employed to examine pigeons' preferences for signaled versus unsignaled delay of reinforcement in which the delay durations ranged from zero to ten seconds. In general, pigeons preferred signaled delay over unsignaled delay especially when a variable-interval 30-second schedule operated in each initial link; when a variableinterval 90-second schedule operated in each initial link, these preferences tended toward indifference or were attenuated. In addition, prior training seemed to exert partial control over behavior. Responding in the terminal link was higher under signaled delay than unsignaled delay in a majority of the cases. Moreover, response rates under signaled delay remained fairly constant whereas responding under unsignaled delay was initially high, but decreased systematically with delay durations as short as 2.5 seconds. These results are consistent with a number of other studies demonstrating the significant role of a signal for impending positive stimuli.Key words: signaled reinforcement delay, unsignaled reinforcement delay, choice behavior, concurrent-chain schedules, key peck, pigeons Over the years, much attention has been allocated to the effects of delayed reinforcement on various aspects of behavior. Early studies by Grice (1948), Perin (1943), and Spence (1947) demonstrated how delayed reinforcement affects the acquisition of new behavior. More recent investigations have examined the effects of delayed reinforcement on maintained response rates (Dews, 1960;Ferster, 1953;Pierce, Hanford, & Zimmerman, 1972;Richards, 1972Richards, , 1981Sizemore & Lattal, 1977 Williams, 1976), and on postdiscrimination gradients of stimulus control (Richards, 1973;Richards & Hittesdorf, 1976Richards & Marcattilio, 1978). Still other inquiries have been concerned with defining a quantitative relationship between responding and the length of the delay interval in choice situations (Chung, 1965;Chung & Herrnstein, 1967;Fantino, 1969 (1970) found that rats learned to escape shock much faster when the negative reinforcement delay interval was signaled compared to when it was unsignaled. Results from two separate studies suggest that responding under conditions of signaled delay decreases as the delay interval increases (Pierce et al., 1972) and that similar results are found under conditions of unsignaled delay except that the latter procedure maintains lower response rates especially at shorter delay durations (as short as 3-sec; Williams, 1976). Recently, Richards (1981) confirmed these general findings in a within-subjects design. In addition, these studies also indicate that responding maintained by unsignaled delay procedures is more variable than responding maintained by signaled delay.Despite the wealth of information this research has contributed, no research to date has examined pigeons' preferences for signaled versus unsignaled reinforcement delay. Such was the purpose of the present investigation. 221 1981, 36,[221][222][223][224][225][226][227][228][229] NUMBER 2 (sE...